The present invention relates to a power amplification circuit to be used for communication devices (cellular phones as an example) or the like. In particular, the invention relates to a power amplification circuit, as well as a communication device using the same, to be used for transmissions that require low-distortion amplification.
In today's and future radio communications systems such as cellular phones and wireless LAN (Local Area Network) systems, digital modulation-demodulation such as QPSK (Quadrature Phase Shift Keying) and OFDM (Orthogonal Frequency Division Multiplex) have become the mainstream, where the power amplification circuit used in these radio communications systems is required to exhibit low-distortion operations. Furthermore, for lower power consumption of battery-driving terminals, the power amplification circuit is required to exhibit high-efficiency operations. As amplification devices to be used for such power amplification circuits, there have been used bipolar transistors and field effect transistors made by using semiconductors such as silicon and gallium arsenide. These bipolar transistors and field effect transistors have a characteristic that the efficiency increases with increasing nearness to the saturation operation. However, amplitude distortion and phase distortion increase in the saturation-operation region, high-efficiency operation and low-distortion operation are in a trade-off relation. To overcome this trade-off, there has been known a technique that negative feedback is applied to a power amplifier.
FIG. 11 shows a common negative feedback power amplifier circuit. In FIG. 11, reference numeral 102 denotes a power amplifier having a gain G2(Pin) dependent on an input signal power Pin and numeral 103 denotes a negative feedback circuit having a feedback quantity β that depends on an impedance thereof, where a negative feedback power amplification circuit 101 is made up by applying a parallel negative feedback to between input and output of the power amplifier 102 with a negative feedback circuit 103.
In this power amplifier 102, as the input signal power Pin increases so that an output signal power Pout approaches saturation, increment of the output signal power Pout becomes smaller than an increment of input signal power Pin, causing the gain G2(Pin) of the amplifier gradually decreases. This gain decrease is a so-called amplitude distortion, which would cause occurrence of a harmonic distortion. Moreover, in the case of amplitude of a modulated-wave input signal having momentary amplitude fluctuations of carriers, the gain would differ among individual momentary carrier amplitudes, causing amplified output signals to have waveform distortions. Resultantly, communication failures such as increases in adjacent-channel power leakage would be incurred.
Meanwhile, a gain G3(Pin) of the negative feedback power amplification circuit 101 resulting from applying a negative feedback to the power amplifier 102 at the feedback quantity β by the negative feedback circuit 103 results in:
                              G3          ⁡                      (            Pin            )                          =                              G2            ⁡                          (              Pin              )                                            1            +                                          G2                ⁡                                  (                  Pin                  )                                            ⁢              β                                                          (                  Eq          .                                          ⁢          1                )            As apparent from Equation 1 above, the negative feedback loop gain G2(Pin)β causes the gain to decrease by a degree corresponding to the factor of 1/(1+G2(Pin)β), as compared with a G2(Pin) that results when no negative feedback is applied. Around the saturation output, an increase in the input signal power Pin causes a decrease in G2(Pin) as well as a decrease in the loop gain. Accordingly, since the above factor increases with increasing input signal power Pin, the decrease in the gain G3(Pin) due to the increase in the input signal power Pin around the saturation output with the negative feedback applied results in a milder one than the G2(Pin) resulting when no negative feedback is applied, as shown in FIG. 12. Therefore, around the saturation output, on condition that the output signal power Pout is of the same, the distortion becomes smaller than when no negative feedback is applied, and on condition that the distortion is of the same level, an operation that is more closer to the saturation can be performed, thus allowing an improvement in efficiency to be obtained.
Further, as can be understood from differentiation of Equation 1 by Pin, it holds that
                                          ⅆ                          G3              ⁡                              (                Pin                )                                                          ⅆ            Pin                          =                                            ⅆ                              G2                ⁡                                  (                  Pin                  )                                                                    ⅆ              Pin                                                          (                              1                +                                                      G2                    ⁡                                          (                      Pin                      )                                                        ⁢                  β                                            )                        2                                              (                  Eq          .                                          ⁢          2                )            Thus, by applying a feedback to the power amplifier 102 according to the factor of 1/(1+G2(Pin)β)2<1, any decrease in the gain of the power amplification circuit in the saturation region can be suppressed. Also, because dG2(Pin)/dPin<0, it follows that dG3(Pin)/dPin<0, so that although the gain decrease is milder than that resulting when no negative feedback is applied, the tendency of gain decreases due to increases in the input signal power Pin is similar to that resulting when no feedback is applied.
Japanese Patent Laid-Open Publication HEI 8-111614 discloses a negative feedback variable-gain power amplification circuit formed from a negative feedback circuit which is provided with a control terminal for a negative feedback quantity so that its gain is controlled by controlling the negative feedback quantity with a control voltage responsive to an input signal power, as shown in FIG. 13. In FIG. 13, reference numeral 171 denotes an amplification-use source-grounded FET (Field Effect Transistor) and 173 denotes a feedback FET, where the source of the feedback FET 173 is connected to the gate of the amplification-use source-grounded FET 171, the gate of the feedback FET 173 is connected to a control terminal 179 via a resistor 177 and the drain of the feedback FET 173 is connected to a control terminal 181. Further, the drain of the amplification-use source-grounded FET 171 is connected to the gate of the feedback FET 173 via a capacitor 175, so that a parallel negative feedback circuit is formed by the feedback FET 173 and the capacitor 175. Also, the control terminals 179, 181 are grounded via bypass capacitors 183, 185, respectively. Further, the gate of the amplification-use source-grounded FET 171 is connected to a signal input terminal 191 via an input matching circuit 187, and the drain of the amplification-use source-grounded FET 171 is connected to a signal output terminal 195 via an output matching circuit 193.
By changing the transconductance of the feedback FET 173 by means of a control voltage V1 applied to the control terminal 179 or a control voltage V2 applied to the control terminal 181, the feedback quantity is changed, by which the gain is controlled. Then, for a large input signal power, the transconductance of the feedback FET 173 is increased by means of the control voltage V1 or V2 to thereby control the gain (more concretely, to decrease the gain) so that the feedback quantity is increased, thus allowing the distortion of the power amplification circuit to be decreased.
However, in the negative feedback power amplification circuit shown in FIG. 11, an attempt to suppress gain decreases to near a saturation operation would require setting the feedback quantity β larger. This would incur, disadvantageously, a decrease in the gain itself as well as a decrease in the power added efficiency (PAE) due to the gain decrease.
In the negative feedback variable-gain power amplification circuit shown in FIG. 13, on the other hand, in which additional control-voltage generating means and input-signal-power detecting means for controlling the feedback quantity are required, the power amplification circuit would be complex in circuitry and moreover hard to downsize, disadvantageously. Further, in order to decrease the distortion for a large input signal power, it becomes necessary to control the gain to a lower one, which would incur a decrease in power added efficiency due to a decrease in gain, as in the case with the negative feedback power amplification circuit of FIG. 11, disadvantageously.